DMLI key terms

The system of decision rights and accountability that determines who may create, change, approve, or retire data assets, and how those decisions are enforced across an organization.

A role accountable for the day-to-day quality, definition, and appropriate use of a specific data domain, acting as the operational bridge between business users and technical teams.

The assignment of ultimate accountability for a data domain to a named individual or role who approves definitions, access rules, and changes affecting that data.

The role responsible for the technical storage, backup, security, and movement of data on behalf of owners, without deciding business meaning or permitted uses.

A formal statement of principles and mandatory rules governing how data is acquired, stored, shared, protected, and disposed of, approved at an authoritative level of the organization.

A documented, reusable rule specifying a required format, value set, or naming convention that data must follow to be considered acceptable for use.

The practice of assigning data to sensitivity or criticality categories so that handling, access, and protection controls can be applied consistently and proportionately.

The mechanisms that decide which users or systems may read, modify, or distribute a data resource, typically through roles, attributes, or explicit permission lists.

A security principle stating that each user or process receives only the minimum data access needed to perform its assigned tasks, for only as long as needed.

Data that can be used, alone or combined with other data, to identify a specific individual, triggering privacy obligations such as notice, consent, and deletion rights.

A documented schedule stating how long each category of data must be kept and when it must be destroyed, driven by legal, regulatory, and operational requirements.

A technique that replaces or obscures sensitive values with realistic substitutes so that non-production users or analysts can work without seeing the original confidential data.

The stages a data asset passes through from creation or acquisition, through storage, use, sharing, archiving, and eventual deletion or destruction.

Structured information that describes other data, such as its name, meaning, format, owner, source, and permitted uses, enabling discovery and correct interpretation.

A reference listing the fields in a dataset or system together with their definitions, data types, allowed values, and relationships to other fields.

An organized inventory of available data assets with searchable descriptions, ownership, lineage, and usage guidance that helps users find and trust suitable data.

A record of the origin and movement of data across systems and transformations, showing how each output was derived from its upstream sources.

The core reference entities shared across business processes and systems, such as customers, products, and locations, whose consistency is essential for reliable operations and reporting.

A controlled set of permitted codes or values used to categorize other data, such as country codes or status values, which must be consistent across systems.

A named aspect used to evaluate fitness of data for use, such as completeness, accuracy, consistency, timeliness, validity, or uniqueness.

A testable condition or threshold that data must satisfy, expressed so that violations can be detected automatically and routed for correction.

The automated examination of a dataset to reveal value distributions, null rates, patterns, duplicates, and anomalies before rules or transformations are designed.

The process of detecting and correcting or removing inaccurate, incomplete, duplicated, or malformed records so that data meets agreed quality expectations.

A processing pattern that extracts data from sources, transforms it to meet target requirements, and loads it into a destination such as a warehouse.

A processing pattern that loads raw source data into a target platform first and then applies transformations there, leveraging the target system's compute power.

A centralized, structured repository optimized for query and reporting that holds integrated, historically retained data drawn from multiple operational systems.

A subset of a data warehouse scoped to a single business area or team, providing focused, easier-to-query data for that group's reporting needs.

A storage repository that holds large volumes of raw or lightly processed data in native formats until a specific analytical need requires it to be structured.

A design approach that organizes data into measurable facts and descriptive dimensions, making it intuitive for business users to query and analyze.

A dimensional design in which a central fact table references surrounding dimension tables directly, producing simple joins and predictable query performance.

A table whose rows record measurable events or transactions, storing numeric measures plus keys that link to the descriptive dimension tables.

A conceptual diagramming method that represents things of interest as entities, their attributes, and the business relationships and cardinalities between them.

The structured design of relational tables to reduce redundancy and update anomalies by separating data into related tables according to dependency rules.

The set of practices and technologies that combine data from multiple sources into a unified, consistent view for operational or analytical use.

The discipline of turning stored data into reports, dashboards, and analyses that support operational monitoring and management decision-making.

A quantified measure tied to a strategic objective, with a defined calculation, target, and owner, used to judge whether performance is on track.

A documented, unambiguous specification of how a measure is calculated, including its numerator, denominator, filters, grain, and source fields.

The presentation of data as charts, graphs, or maps chosen to reveal patterns and comparisons clearly and honestly to a defined audience.

The application of moral principles to how data is collected, used, and shared, weighing fairness, transparency, consent, and potential harm to affected people.

Systematic unfairness in automated outputs that arises from skewed training data, flawed proxies, or design choices, producing unequal outcomes for groups of people.